Beckett and Proudfoot • NORTHERN SAW-WHET OWL MIGRATION 
533 
Predicted adult : juvenile ratio 
1 1 0.01-0.38 
\^\ 0.39-0.64 
^ 0.65-0.82 
0 83 - 103 
1.04-1.38 
1.39-9.33 
Banding station 
with >50 banding 
events 
FIG. 1C. Continued. 
The wide directionality distribution of migrants in 
'he Great Lakes Basin illustrates that owls traveling 
through this region have many directional options, 
supporting our conclusions of high route-fidelity in 
this region (Table 1). 
Overall observed directionality may be influ¬ 
enced by the distribution of banding stations 
across eastern North America. For example, owls 
migrating southeast from western Pennsylvania 
would be more represented in the data base than 
owls migrating southwest due to the location and 
number of ‘downstream' stations (Fig. 1). Station 
location biases may explain the significant 
difference in directionality between owls recap¬ 
tured >100 versus >500 km from banding 
location. However, such biases do not obscure 
our general conclusion that Northern Saw-whet 
Owls are migrating predominately southward. 
Age-differentiated Migration .—Juveniles were 
banded more frequently than adults, and this is 
consistent with local studies (Weir et al. 1980, 
Stock et al. 2006, Brittain et al. 2009). Banders in 
eastern North America report a striking increase in 
the proportion of juveniles banded in irruption 
years (Brinker et al, 1997, Paxton and Watts 2000, 
Whalen and Walts 2002). Our results support this 
finding. The percentage ot adult Northern Saw- 
whet Owls was lowest in 1999 and 2007, two well- 
recognized irruption years, and highest in years 
immediately following these irruptions (Table 2). 
This is consistent with local findings as well 
(Paxton and Watts 2000). The discrepancy in 
proportions of juveniles is likely due to the cyclical 
prey base causing high reproductive success in 
irruption years followed by poor success the 
following year (Cheveau et al. 2004). Also, many 
adults in post-irruption years are returning second- 
year birds that hatched the previous year. 
Our results did not reveal unequivocal evidence 
of age-differentiated migration based on latitude 
